Sound Compatibility: "Normal Writing Vs. Direct Writing"

[TempesT]

Under the options menu in R1Q2, for sound compatibility, it says "normal writing" and "direct writing". It usually says "max compatibility" and "max performance" instead. What I'm wondering is, what setting out of "normal writing" or "direct writing" would have the best sound quality? I'm looking for the highest quality sound, so out of those choices, what would bring the highest quality audio?

[Jay Dolan]

Those two options have never had anything to do with sound quality; they are for performance. With normal writing / max compatibility, Quake writes the mixed audio stream to a buffer in memory, and informs the sound card of the length of the buffer at each frame. With direct writing / max performance, Quake maps the audio driver's memory buffer, and writes to it directly. This saves the audio driver from having to copy the user buffer at each frame. In 1997, that was sometimes a 1-3fps boost. Today, it's meaningless. I doubt you'll notice a performance difference with either, but feel free to timedemo it.

[TempesT]

Thank you for the detailed explanation Jay. I did further researching between the two settings, and discovered that if you have a lower end soundcard, that you should choose normal writing/max compatibility. But, if you have a decent sound card, that you should use direct writing/max performance so you can take advantage of the DirectSound primary buffer.


Incase anyone wants to know, the cvar that controls this setting is "s_primary". A value of 0 will use "normal writing/max compatibility", and a value of 1 will use "direct writing/max performance".